Robust Adaptive Beamforming Algorithm Research

The performance is distorted badly in practice by array error. Hence, it is very important in theory and application to study on robust beamforming. we first review the classical models and algorithms in robust adaptive beamforming, and work on the both narrow band and Ultra-wideband (UWB) beamformer, and then propose some algorithms.Since the performance of beamformer methods suffers from severe degradation if there is a small mismatch in the steering vector. Base on the narrow band signals, Our method belongs to the class of diagonal loading, but the optimal amount of diagonal loading level can be precisely calculated based on the uncertainty set of the steering vector, and a beamformer base on the spare constraint is proposed, which can not only reduce the sidelobe level, but also increase robustness against the mismatch problem caused by imprecise knowledge of the desired source. The UWB beamformer is studied base on the frequency domain, since the analog delay is hard to apply in practice. So a method based on the conventional MVDR of muti-points frequencies is proposed, which can not only overcome the degradation of the capability, but constrain the interferences.In the situation of a few snapshots, the eigenvalue and eigenvector corresponding to the expectation signal and noise are concerned with adaptive beamforming, which accordingly causes the adaptive beamforming distortion. Because of that, an adaptive method is presented by analyzing the eigenvalues of sampling covariance matrix in order to improve the performance and robustness of beamforming. Computer simulations have demonstrated the good performance using all the presented methods.